Cohort Analysis: Why Every SaaS Company Needs To Be Doing It

Every SaaS company not only wants someone to sign up for their app or software but also to have them continuously come back for repeat usage and recurring subscription payments. In order to tell how you’re progressing on these goals, you will need to look further than just basic metrics such as total subscriber counts or total monthly active users. In order to get the insights that you’re looking for, you will need to use cohort analysis to dive deeper into your user base.

What is Cohort Analysis?

Cohort analysis is a behavioral analysis that takes the data from a given user data set and, instead of looking at all users as one single unit, it breaks them up into smaller segmented groups for analysis. These groups, or cohorts, tend to share common characteristics or experiences within a defined timeframe. By using a Cohort Analysis you’re able to analyze patterns clearly across the lifespan of a customer (or a user) instead of just blindly looking through aggregate data of all customers without accounting for the natural lifecycle of a customer. With these reports, you’ll be able to truly understand your customer and tailor your services to those specific groups/cohorts. While it’s important to know how many downloads your app got in total, the usefulness of such a broad high-level metric can be limited. With cohort analysis, you’re able to spot patterns at multiple points in the customer lifecycle and understand their behavioral changes, which then can help guide you in product decisions and development to make sure your product suits the needs of your users.

Here is an example from HubSpot of what a cohort analysis looks like:

Cohort Analysis

Now that we have gone over the basics, let’s see how you can increase retention when you use cohort analysis for reporting. 

Why Is Cohort Analysis Useful?

The information you get from a  cohort analysis is valuable because of the specificity of the information it provides. It especially helps companies find answers to targeted questions by analyzing relevant data.

Here are a couple of ways it can help SaaS companies analyze and answer the tough questions about their users/customers.

  • Learn how user behavior affects your business. By analyzing behaviors of cohorts, you can understand what actions people did or didn’t take in a specific timeframe and how that translates into business metrics, such as retention and acquisition. 
  • Understand customer churn. Gathering your data gives you the ability to assess a hypothesis such as whether sign-ups related to a specific promotion lead to greater or lesser churn.
  • Calculate average customer lifespan. How long do you typically hold on to most users?
  • Calculate customer lifetime value. With a cohort analysis, you can go beyond a basic high-level customer lifetime value calculation to instead take a look at specific timeframes — for example, grouping by signup month — to see which specific months provided higher value. Going further than that, you can then group them by size, segment, and time to understand further which channels and campaigns lead to the best customer lifetime value.
  • Optimize your conversion funnel. Being able to compare customers who are engaged in various ways at given times in your sales process allows you to understand how your user experience in your marketing efforts translates into new business opportunities.
  • Create more effective customer engagement. As you see patterns in how various cohorts engage with your company, website, and product, you can take steps that will encourage all customers to take various actions more efficiently.

Cohort Data

Performing a cohort analysis differs from company to company. So to set up an analysis that works for you, you’ll need to sit down and think about what questions you’re trying to answer. You will need to select the following criteria from wherever you get your data:

  1. The characteristics of your cohort (what defines this group)
  2. An inclusion metric (the action that precipitated inclusion in the group)
  3. A return metric (the thing you want to know about them)

In order to track and analyze how users behave over time or how the same behavior differs between different groups,  cohort analysis helps you compare these people by the way/time they were acquired or by the retention of those users over time. 

Example 1

Let’s say you’ve developed a mobile game app and you’re trying to analyze if iOS users have been more or less profitable than Android users over the last quarter. Since you’ve spent the same amount of money promoting and marketing to both platforms, you decide you want to measure how valuable users are on each platform by comparing the average revenue per user (ARPU) between users on iOS devices and Android devices.

In this example, the characteristics of your cohort, like I mentioned above, are defined by the mobile operating system each user has — iOS or Android. The inclusion metric in this instance would be active users over the last quarter. Lastly, the return metric would be the average revenue per user.

Once we have this data, what can we figure out? In the example, let’s say the inclusion tells you the iOS cohort has 400,000 total users and the Android one has 500,000 total users. It also indicates that the iOS cohort has 200,000 daily active users over the last quarter and the Android cohort has 250,000 daily active users during the same time period. But the return metric shows that the iOS cohort has an ARPU of $3 and the Android cohort has an ARPU of $2.

Now, what can we gather from this information that will help this business? Well, you might conclude that, while your iOS daily active user base currently is a bit smaller than your Android daily active user base, they nevertheless are 50% more profitable than Android users. With this information, it might make the most sense to allocate more marketing resources to reach and promote to iOS users next quarter.

Example 2

Now let’s look at another analysis. Say you’ve got a cloud-based time-tracking software and you want to be able to compare the retention rates of the customers you acquire from two distinct marketing campaigns: those who signed up through a HubSpot email drip campaign in June of last year, and those who signed up from a Google Ads Campaign in July of last year.

The characteristics of your cohorts are the two marketing campaigns attributed to the new customer — email or AdWords. The inclusion metric is the action of signing up. And the return metric would be the number of remaining customers as of this current month.

For this example, let’s say the email drip cohort had 200 initial customers, while the Google Ads had 300. Now the return metric for email shows 100 currently-remaining customers, while Google Ads shows 250 currently-remaining customers. Thus, the retention rates are 50% for those who sign up through the email campaign and 83% through your Google Ads.

I think the answer is pretty obvious with this one. You can see that Google Ads generates a significantly higher amount of long-term retained customers than your email campaigns. So it would probably be best to spend more money on more Google Ads in the future.

The Answers Are In The Data

You’re probably already collecting user data at your company today, so the data is all there. It’s just a matter of breaking up your data into groups to understand behaviors and have actionable next steps to your marketing efforts. You’d be surprised by the answers you can get through a cohort analysis.

This post was originally published August 2020 and has been updated for accuracy.